i2p.www/www.i2p2/pages/meeting136.html

{% extends "_layout.html" %}
{% block title %}I2P Development Meeting 136{% endblock %}
{% block content %}<h3>I2P dev meeting, April 5, 2005</h3>
<div class="irclog">
<p>14:34 &lt;@jrandom&gt; 0) hi</p>
<p>14:34 &lt;@jrandom&gt; 1) 0.5.0.5</p>
<p>14:34 &lt;@jrandom&gt; 2) Bayesian peer profiling</p>
<p>14:34 &lt;@jrandom&gt; 3) Q</p>
<p>14:34 &lt;@jrandom&gt; 4) ???</p>
<p>14:35 &lt;@jrandom&gt; 0) hi</p>
<p>14:35  * jrandom waves</p>
<p>14:35  * smeghead outsources his todo list to a parallel universe</p>
<p>14:35 &lt;@jrandom&gt; weekly status notes posted up @ http://dev.i2p.net/pipermail/i2p/2005-April/000675.html</p>
<p>14:36 &lt;@jrandom&gt; might as well jump on in to 1) 0.5.0.5</p>
<p>14:36 &lt;+ant&gt; * Connelly waves</p>
<p>14:37 &lt;+protokol&gt; high everyone</p>
<p>14:37 &lt;@jrandom&gt; as mentioned in the status notes (and the current history.txt), we've tracked down some very long lasting netDb bugs</p>
<p>14:37 &lt;@jrandom&gt; in the past, we've been able to fudge it, but 0.5.0.5 forced us to start doing things "right", which is why its been biting us now</p>
<p>14:39 &lt;@jrandom&gt; i expect we'll have a new release sometime tomorrow, so keep an eye out for the update link on your router console :)</p>
<p>14:39 &lt;+protokol&gt; yey</p>
<p>14:39 &lt;@jrandom&gt; actually, thats about all i have on that at the moment - anyone else have anything to add wrt 0.5.0.5?</p>
<p>14:40 &lt;+protokol&gt; nope</p>
<p>14:41 &lt;@jrandom&gt; ok, moving on to 2) Bayesian peer profiling</p>
<p>14:41 &lt;@jrandom&gt; ah, damn, bla dropped off the channel a few mins back</p>
<p>14:42 &lt;@jrandom&gt; well, anyway, I just wanted to point people out at bla's work exploring some more robust profiling techniques</p>
<p>14:42 &lt;+protokol&gt; postponing 2?</p>
<p>14:43 &lt;@jrandom&gt; check out the forum post and the link to theland.i2p for more info, and bounce bla your thoughts :)</p>
<p>14:44 &lt;@jrandom&gt; ok, movin' on to 3) Q</p>
<p>14:44 &lt;@jrandom&gt; aum: you up?</p>
<p>14:44 &lt;@jrandom&gt; hmm, doesnt look like it</p>
<p>14:45 &lt;@jrandom&gt; ok, lots of progress on the Q front, more details for getting involved in alpha testing up @ http://aum.i2p/q/ </p>
<p>14:45 &lt;@jrandom&gt; i'm sure we'll hear more on the list when there's an update available</p>
<p>14:46 &lt;+ant&gt; &lt;Connelly&gt; Q works for me for retrieving content</p>
<p>14:46 &lt;@jrandom&gt; yeah, its been working great for me as well, a few bumps here and there, but quite promising</p>
<p>14:47 &lt;+ant&gt; &lt;Connelly&gt; my Q server stored 2 small items, then got stuck at 100% cpu usage until i killed it</p>
<p>14:47 &lt; zzz&gt; for those who haven't seen it check out my q front end http://flock.i2p/cgi-bin/q</p>
<p>14:47 &lt;@jrandom&gt; zzz: that is quite kickass</p>
<p>14:48  * jrandom forgot the url to that when writing up the status notes (d'oh)</p>
<p>14:50 &lt;@jrandom&gt; ok, anything else on 3) Q?  or shall we move on to 4) ???</p>
<p>14:50  * jrandom considers us moved</p>
<p>14:51 &lt;@jrandom&gt; anyone have anything else they'd like to bring up for the meeting?</p>
<p>14:51 &lt;+ant&gt; &lt;Connelly&gt; i've coding an http/html filter for i2p</p>
<p>14:51 &lt;+protokol&gt; yes</p>
<p>14:51 &lt;+protokol&gt; ian clarke is a troll on slashdot</p>
<p>14:51 &lt;+ant&gt; &lt;Connelly&gt; been coding</p>
<p>14:51 &lt;+ant&gt; &lt;Connelly&gt; should be more safe than freenet's html filterer</p>
<p>14:51 &lt;+ant&gt; &lt;Connelly&gt; if i run out of time i'll just incorporate freenet's filterer</p>
<p>14:51 &lt;@jrandom&gt; cool Connelly, how is it coming along?</p>
<p>14:52 &lt;@jrandom&gt; protokol: and you're a troll in #i2p ;)</p>
<p>14:52 &lt;+ant&gt; &lt;Connelly&gt; so in the end we should have an html filterer for i2p</p>
<p>14:52 &lt;+ant&gt; &lt;Connelly&gt; got html filtering done, now working on css, still haven't looked at header filtering</p>
<p>14:53 &lt;+ant&gt; &lt;Connelly&gt; it's very paranoid :)</p>
<p>14:53 &lt;@jrandom&gt; bitchin</p>
<p>14:53 &lt;+protokol&gt; whitelist?</p>
<p>14:53 &lt;@duck&gt; does it let anything trough at all?</p>
<p>14:53 &lt;+ant&gt; &lt;Connelly&gt; yeah</p>
<p>14:53 &lt;+protokol&gt; if so, what is currently disallowed</p>
<p>14:53 &lt;+protokol&gt; (of any importance)</p>
<p>14:55 &lt;+ant&gt; &lt;Connelly&gt; disallowed of significance: frames and iframes, scripting, optgroup</p>
<p>14:55 &lt;+ant&gt; &lt;Connelly&gt; meta</p>
<p>14:55 &lt;+ant&gt; &lt;Connelly&gt; embedded objects</p>
<p>14:56 &lt;@jrandom&gt; neat.  i'm looking forward to seeing how things progress - any eta on when we could try rigging it up with the eepproxy?</p>
<p>14:56 &lt;+ant&gt; &lt;Connelly&gt; i'll probably have an alpha in 1-2 weeks</p>
<p>14:57 &lt;+ant&gt; &lt;Connelly&gt; so we can test out how it works</p>
<p>14:57 &lt; jrandom2p&gt; kickass</p>
<p>14:58 &lt;+ant&gt; &lt;Connelly&gt; it allows forms, cookies, content caching but those can be turned off in 'paranoid' mode</p>
<p>14:58 &lt;+protokol&gt; why frames and iframes? can you just not block connections to non-i2p sites from them?</p>
<p>14:59 &lt;+ant&gt; &lt;Connelly&gt; it has a cgiproxy like url navigator bar at the top</p>
<p>14:59 &lt;+ant&gt; &lt;BS314159&gt; I suspect the hard thing would be blocking frames between different eepsites</p>
<p>14:59 &lt;+ant&gt; &lt;Connelly&gt; i don't want that hijacked</p>
<p>14:59 &lt;+protokol&gt; i mean can you just block connections</p>
<p>14:59 &lt;+ant&gt; &lt;Connelly&gt; could make it like freenet's proxy where you just enter a url at the beginning</p>
<p>14:59 &lt;+protokol&gt; yeah, frames can rock</p>
<p>14:59 &lt;+ant&gt; &lt;Connelly&gt; and can't enter urls once you start browsing</p>
<p>14:59 &lt; jrandom2p&gt; frames kill kittens</p>
<p>15:00 &lt;+ant&gt; &lt;BS314159&gt; this has to be the oldest framewar ever.  excuse me, flamewar</p>
<p>15:00 &lt; jrandom2p&gt; heh</p>
<p>15:00 &lt;+protokol&gt; i said "can" rock</p>
<p>15:00 &lt;+ant&gt; &lt;BS314159&gt; what we need is our own browser</p>
<p>15:00 &lt;@jrandom&gt; and flying ponies</p>
<p>15:01 &lt;@jrandom&gt; *cough*</p>
<p>15:01 &lt;+ant&gt; &lt;Connelly&gt; i'd prefer an F-16 to a pony</p>
<p>15:01 &lt; Teal`c__&gt; can i have a girl ?</p>
<p>15:01 &lt;+ant&gt; &lt;Connelly&gt; i'll make an option for enabling frames</p>
<p>15:01 &lt;+protokol&gt; Teal`c__: no</p>
<p>15:02 &lt;+ant&gt; &lt;BS314159&gt; Is there a functioning I2P inproxy? bolas.mine.nu appears to be dead.</p>
<p>15:02 &lt;+protokol&gt; from other eepsites, right?</p>
<p>15:02 &lt;@jrandom&gt; BS314159: http://i2p.mine.nu/</p>
<p>15:02 &lt;+protokol&gt; i2p.mine.nu</p>
<p>15:02 &lt; frosk&gt; i2p.mine.nu</p>
<p>15:02 &lt;+ant&gt; &lt;BS314159&gt; thanks</p>
<p>15:02 &lt;+ant&gt; &lt;BS314159&gt; frames are safe if they're inside one eepsite.  frames are safe if all content is static</p>
<p>15:03 &lt;+ant&gt; &lt;BS314159&gt; the only danger is if there's a form in one of the frames, since you might submit information to the wrong party</p>
<p>15:04 &lt;@jrandom&gt; eh, i'm of the opinion the filter should only support what we *need* (and know is safe), and let actual end user demands expand functionality, rather than preemptively assume people will want some things</p>
<p>15:04 &lt;+ant&gt; &lt;BS314159&gt; wise</p>
<p>15:06 &lt;@jrandom&gt; ok, anyone else have anything for the meeting?</p>
<p>15:06 &lt; Teal`c__&gt; sorry didn't know a meeting was on</p>
<p>15:07 &lt;@jrandom&gt; heh no worry, you'll be immortalized in the meeting logs ;)</p>
<p>15:07 &lt;@jrandom&gt; speaking of which</p>
<p>15:07  * jrandom winds up</p>
<p>15:07  * jrandom *baf*s the meeting closed</p>
</div>
{% endblock %}